Flower AI and Vana have joined forces to launch Collective-1, a distributed large language model breaking free from traditional data center limitations. By leveraging a global network of GPUs, this innovative project bypasses centralized infrastructure dependence and utilizes private user data from platforms like X, Reddit, and Telegram. While currently boasting 7 billion parameters, Flower AI is aiming for a 30 billion parameter model release in the near future, with a larger version reaching nearly 100 billion parameters within the year. The company’s efforts to democratize AI are further evident through their collaborative development of Photon, an efficient tool designed for distributed training. This project empowers smaller businesses and resource-limited nations to participate in foundational model development. Vana provides user control over data usage with its authorization tools, allowing users to decide on how their data is utilized and reap the benefits thereof. Experts believe this approach will reshape the AI industry towards a more open and decentralized future.